Truss Design failing members

Question

I need to design truss steel structure 50x80 mts as per IS code. Solar panel is mounted on truss structure acting as a roof. I have analysed the structure. Is this correct way ? Lot of members fail. Please point out the mistake have done

Surojit Ghosh · Answer

I have reviewed your model. Few things I want to mention-- 
 
1. You have specified all the roof members as truss member for which there are several zero stiffness warning messages in the output. You need to modify the member specification from MEMBER TRUSS to MEMBER RELEASE at both end with 0.5% fixity. For more details on this, refer to the following post-- 
 communities.bentley.com/.../staad-pro-instability-and-zero-stiffness-faq.aspx 
 
2. You have used IS800:1984 code (working stress design method) but used factored load combination for the design. As per cl.3.4.2.1, unfactored load combination cases should be considered for design. You can find that almost for all the failed members, the critical load case is factored load combination cases (starting from 201). You need to review the combination cases defined in your model. 
 
3. The critical clause for the failure is either combined interaction (7.1.1) or slenderness. Also there are several warning messages mentioned in the design output as-- 
 
**WARNING- THE AXIAL COMPRESSIVE STRESS EXCEEDS THE EULER STRESS FOR MEMBER 2096 
 EQN. 7.1.1(A) CANNOT BE CHECKED PROPERLY. 
 
If you see the output file you would notice that the actual axial stress (fa) is greater than the elastic critical stress. 
 
Now, in the Equation no. 7.1.1(A) of IS800:1984 , the term (1-fa/fcc) is in the denominators. So, as “fa’ is greater than fccz, the term (1-fa/fccz) becomes negative which is not permitted by the code. So, when STAAD encounters this condition, it produces the warning message, and sets the value of the ratio for the associated bending terms in cl.7.1.1(A) to a very large number. In this situation, the member had already buckled and computing the whole Interaction ratio value is unimportant and hence the ratio thus reported is just a large value. 
 
You may want to use a larger cross section so that "fa" becomes smaller, or use one with a smaller KL/r value so that fccy and/or fccz become larger than the axial stress.